ODBC is actually pretty effecient, at least the one from MiniSoft which is
where my experience is.  The server parses down the data you need and then
passes it down, as opposed to passing everything down and having you
program figure out what to use.  If you want to write custom servers then
you could get one of the various middleware packages available and write
your own servers to maximize performance.

I personally use ODBC/32 and MiddleMan from MiniSoft, and then when
required we write our own server for some really hard core Omnidex stuff
and use the MiddleMan STREAMS access server.  We get really phenomonal
performance considering the architecture.






"Holger Schr?rs" <[log in to unmask]> on 01/20/2000 04:51:14
PM

Please respond to "Holger Schr?rs" <[log in to unmask]>



 To:      [log in to unmask]

 cc:      (bcc: Shawn Gordon/IS/FHM/FHS)



 Subject: Data-Exchange with HP3000







Hi Anybody !   ;-))

Now I'm working only a half year on a HP3000 - 979.
I know that an ODBC-Interface could make it possible
to use Turboimage/Omnidex-Databases directly on a
PC (inside Windows ect.).

Did You know any other possibility?
A faster, one with less CPU-usage?
The ODBC-Stuff uses too much CPU-capacity.

Thanks for an answer!!

Holger Schr




örs
Germany